In June while we were still under 2300, I was forecasting a series called "3000 BY HALLOWEEN". I have gave up on it several times because I didn't trust my work enough, simply bc my short-medium engine never agreed with my medium long engine.
I realize now the error in that logic. The short-medium engine can only agree with medium long engine when they are forecasting the same window, as in the next 35 days. So this is the "raw base case". I expect the 2593-2600 correction after NFP. This move accomplishes the rejection at the channel and the 2593-2600 retrace at the same time. This leaves only one realistic curve to 3000 and blue route in chart above is that "generic route".
I will post DRAFT 10-2 when I have it.